Method, apparatus and computer program product for classification of measured data from multi-core optical fiber

There is provided a method, an apparatus (5) and computer program product for the classification of measured data from a multi-core optical fiber (1), wherein even with event points which are difficult to distinguish or with identical event points and when there are deviations in distances between event points, connection losses and the like for all event points can be classified. Measured waveform data measured using an OTDR apparatus(2) is read in (S1). Then level differences with connection losses which coincide or exceed a predetermined threshold value are identified as event points (E11~E13, E21~E23, E31~E32, E41~E43), and the event information related to the event points is obtained (S2). Next the event information related to the head core optical fiber (C1) is registered in a master file (S3). Then an interval distance between adjacent event points registered in the master file is obtained (S4). After this, event points which exist within a predetermined allowable range of the event points registered in the master file are searched out from each event point in the object core optical fiber, and determined to be identical event points (S5). Then if the event distance of the object core optical fiber is less than the event distance registered in the master file, the event information registered in the master file is updated using the event information for the object core optical fiber (S6). After this, of the event points included in the object core optical fiber, all the event information related to event points which are not identical event points is registered in the master file (S7). Then the processing described above (S4-S7) are executed for all core optical fibers (C1, C2, C3, C4, ..., ) (S8, S9). Finally connection losses and the like are classified for each event point (S10).